There has been plenty made of the potential for artificial intelligence (AI) to transform businesses. But where do you even start?

The number one rule for getting started with AI is to keep it simple, said analytics software firm SAS’ COO Oliver Schabenberger at an event in Sydney last week.

“Do not boil the ocean,” he said. “If your organisation is not a digital native start small, identify one to three projects in a core competency of the business.”

Next, think about what your business does.

“The project should be industry-specific,” Schabenberger continued.

“Someone else will do a much better job at building these non-industry-specific solutions than you could ever do it yourself in house.”

Industry-specific solutions will be more likely to deliver a ROI, he explained, which in turn means they will be more likely to be signed off for funding.

And although the prospect of introducing AI to a business for the first time might be an intimidating one, it is important to keep the team small for the pilot.

This will make communication more effective and again help to convince the rest of the business that the project can deliver a ROI.

Making it mainstream

As well as getting first-timers started, Schabenberger also highlighted the ways in which businesses can start to get true value from AI and automation.

There is currently a gap between the potential of AI and its realised value.

“We have to solve how to integrate,” Schabenberger said. “How to integrate this [data-driven automation] into business processes and operations.”

“How to take AI from a science project to creating value for the organisation.”

Central to this is the idea of ‘democratising’ AI and giving everyone within an organisation access to the powerful insights available, described by Schabenberger as “analytics for everyone, everywhere”.

But this is not possible in the current environment.

“We have a communication gap,” he said. “A communication barrier between the data scientists and the business.

“There’s real difficulty taking analytics from the science project to operational excellence.”

Overcoming this requires “access to resources”, he explained.

“The resource here is data, the engine is analytics.

“That’s how we create value.

“The value of data without analytics is not realised.”